Open access Original research BMJ Open: first published as 10.1136/bmjopen-2020-038772 on 29 September 2020. Downloaded from Patterns and predictors of mortality in a semi-­urban population-based­ cohort in Sri Lanka: findings from the Ragama Health Study

Anuradhani Kasturiratne ‍ ‍ ,1 Dileepa Senajith Ediriweera,2 Shamila Thivanshi De Silva,3 Madunil Anuk Niriella,3 Uthuru Beddage Thulani,1 Arunasalam Pathmeswaran ‍ ‍ ,1 Anuradha Supun Dassanayake,4 Arjuna Priyadarsin De Silva,3 Sureka Chackrewarthy,5 Udaya Ranawaka ‍ ‍ ,3 Norihiro Kato,6 Ananda Rajitha Wickremasinghe,1 Hithanadura Janaka de Silva3

To cite: Kasturiratne A, ABSTRACT Strengths and limitations of this study Ediriweera DS, De Silva ST, Objective To describe patterns and predictors of mortality et al. Patterns and in a semi-urban­ population in Sri Lanka. ►► This is a population-based­ prospective cohort study predictors of mortality in Design A prospective population-­based cohort study. a semi-urban­ population-­ which provides high quality epidemiological evi- Setting Ragama Medical Officer of Health area in the based cohort in Sri Lanka: dence on temporal associations. Gampaha district, Sri Lanka. findings from the Ragama ►► We used national vital registration data with a high Participants Adults between 35 and 64 years of age Health Study. BMJ Open coverage and completeness (99.9%), but they may were recruited using an age stratified random sampling 2020;10:e038772. doi:10.1136/ have inaccuracies in documenting the cause of technique in 2007. bmjopen-2020-038772 death. Measures At baseline, we recorded socio-­demographic, ►► Prepublication history for ►► Loss to follow-­up of the original cohort (8.7%) is a lifestyle, anthropometric, biochemical and clinical data this paper is available online. limitation of this study. of the participants. Over 10 years, we obtained the To view these files, please visit ►► Relationships observed may be distorted by con- cause and date of death from the death registration the journal online (http://​dx.​doi.​ founding, but multivariable analysis that have been documents of deceased participants. We determined org/10.​ ​1136/bmjopen-​ ​2020-​ carried out is likely to have minimised this. 038772). the survival probability of the cohort over 10 years and estimated Hazard ratios (HRs) for all-­cause mortality Received 28 March 2020 (ACM), cardiovascular mortality (CVM) and cancer-­related http://bmjopen.bmj.com/ Revised 13 August 2020 mortality (CRM) using Cox’s proportional hazards model. easiest to measure due to it being an unambig- 1 Accepted 14 August 2020 We also estimated the survival probabilities for men and uous endpoint. Due to rapid socio-economic­ women in each 10-­year age group and standardised development and improved living conditions, mortality ratio relative to the source population. significant reductions in adult mortality have Results There were 169 deaths over 10 years with been observed in all regions of the world in standardised mortality rates of 5.3 and 2.4 per 1000 years the last three decades. In both high-income­ of follow-up­ for men and women, respectively. Independent and middle-income­ countries the the most predictors of: ACM were older age, lower income, smoking common cause of death is cardiovascular on October 1, 2021 by guest. Protected copyright. and diabetes mellitus while gender, education, occupation, disease. South Asia, with a population of harmful alcohol use, waist circumference and hypertension 1.9 billion, records over 12 million deaths were not; CVM were older age, lower income, smoking, 2 diabetes and hypertension while gender and harmful alcohol annually and the majority of these deaths are use were not; CRM was older age while gender, smoking and due to cardiovascular causes which are linked diabetes were not. Those engaged in clerical and technical to intermediate metabolic conditions such as, 3 occupations or unemployed had a lower risk of CRM as diabetes mellitus and hypertension. It has © Author(s) (or their compared with those engaged in elementary occupations. been documented that people of South Asian employer(s)) 2020. Re-­use Conclusions Older age, lower income, smoking, diabetes origin develop diabetes mellitus at a younger permitted under CC BY-­NC. No and hypertension strongly predict mortality in this cohort. commercial re-­use. See rights age and at lower levels of metabolic risk when 4 and permissions. Published by Addressing the identified modifiable predictors through compared with Caucasians. South Asians BMJ. behavioural modification will improve longevity in similar who make up a considerable proportion of For numbered affiliations see populations. the migrant population in high-income­ coun- end of article. tries have also shown a higher risk of diabetes INTRODUCTION at lower metabolic risk levels.5 6 Mortality Correspondence to Mortality is one of the strongest indicators due to cancer is estimated to be about Dr Anuradhani Kasturiratne; of the health status in a population and the anuradhani@​ ​kln.ac.​ ​lk 1.3 million deaths in South-East­ Asia region7

Kasturiratne A, et al. BMJ Open 2020;10:e038772. doi:10.1136/bmjopen-2020-038772 1 Open access BMJ Open: first published as 10.1136/bmjopen-2020-038772 on 29 September 2020. Downloaded from and accounts for about 10% of mortality in the region.8 years for women.14 The district is divided into 16 MOH All cause, cardiovascular and cancer mortality in Asia areas to provide population based preventive healthcare have shown strong links to socio-­economic disadvan- services.16 The Ragama MOH area, which has semi-urban­ tage.9 However, only a few studies have been conducted characteristics and a population of 81 687 living in an area to describe the patterns and predictors of mortality in of 21 km2, is the smallest of these and is the field practice South Asian populations and even fewer are based on area of the Faculty of Medicine, University of Kelaniya.17 population-­based cohorts.10 At the commencement of the study in 2007, 3012 adults Well organised and functioning vital registration systems between 35 and 64 years of age living in the Ragama MOH provide high quality and timely mortality data that are of area were recruited using age stratified random sampling great value for epidemiological purposes. At present, the from the electoral register (sampling frame) conforming rate of death registration in Sri Lanka exceeds 99% and on to the ratio of 1:2:2 from each of the three 10-year­ age registration, a death certificate is issued. The death certif- groups (35 to 44, 45 to 54 and 55 to 64 years, respec- icate includes the identification data of the deceased, the tively).18 Participant recruitment was done between July time, date, place and the medical cause of death and the 2007 and March 2008. At the baseline assessment 2985 designation and credentials of the officer certifying the of the recruited participants had complete data and were death; it is required for the disposal of the deceased.11 followed-­up as the Ragama Health Study (RHS) cohort. Although data from the vital registration system and We included all the participants of the RHS cohort in this hospital statistics provide evidence on the importance analysis. of cardiovascular disease as the leading cause of death across South Asia,12 13 there is limited epidemiological Data and data sources information on the patterns and predictors of mortality After providing informed written consent at recruitment, from population-based­ studies conducted in this region. the study participants underwent baseline screening that This dearth of information hampers creating the aware- included anthropometric, clinical and biochemical assess- ness it deserves for better decision-­making and resource ments and their socio-demographic,­ lifestyle, medical and allocation for appropriate public health and preventive health-related­ information were obtained through an measures for control of behavioural and metabolic risk interviewer-administered­ questionnaire developed in the factors of cardiovascular disease. Results from studies local languages. Medical history provided by the partici- conducted outside South Asia have a limited impact on pants was verified with medical records by the medically attracting the attention of local decision-makers­ and qualified research staff. The assessment covered a wide health administrators. range of risk factors attributed to cardiovascular diseases To address this dearth of information and the resultant including biochemical tests for assay of blood glucose and gaps in practice, we conducted an analysis to describe lipids, serum insulin, glycated haemoglobin and alanine the patterns of all-cause­ mortality (ACM), cardiovascular transaminase levels. In 2010 and 2014, two follow-up­ mortality (CVM) and cancer-related­ mortality (CRM) assessments of the cohort were conducted using similar http://bmjopen.bmj.com/ and their predictors in a population-based­ cohort in Sri methods. In 2014, 2148 (72%) of the 2985 original partic- Lanka, using baseline data of the cohort and mortality ipants attended follow-­up. The rest of the participants data reported by the vital registration system. were followed-up­ in the community by research staff and their vital status was determined. If any participant had died, a copy of his/her death certificate was obtained and METHODS details were recorded. From July 2017 to November 2018, Study design we conducted a population based follow-­up of the RHS

We prospectively followed-up­ participants of a population cohort at household level by contacting all households by on October 1, 2021 by guest. Protected copyright. based cohort study over a period of 10 years to correlate telephone or by post, followed by home visits if a death their baseline characteristics and mortality outcomes. was reported. The follow-up­ visits were conducted by trained research assistants with a medical/health sciences Study setting and population background, who met family members of the deceased Our cohort lives in the Ragama Medical Officer of participants to collect information on their cause of Health (MOH) area in the Gampaha district of Sri death. They obtained a copy of the death certificate to Lanka. Gampaha district is the second most populous correctly ascertain the date and the cause of death. If the district of Sri Lanka with a population of approximately death certificate was not available with the family, more 2.3 million.14 Sri Lankans enjoy universal free education preliminary hospital or field death declaration docu- and free healthcare facilities at the point of delivery. ments were perused to ascertain the date and the cause Universal franchise was introduced in 1931 and all citi- of death. The vital status of 2724 (91.3%) participants was zens above 18 years of age are included in the electoral obtained. The database was updated with the outcome register since 1959.15 Gampaha district with mostly rural data obtained during the follow-up­ assessment after a and semi-urban­ characteristics, records the country’s cleaning and verification process. At each stage of this highest overall adult literacy rate of 98.5% and an overall prospective cohort study, participants who did not partic- life expectancy at birth of 73.2 years for men and 79.9 ipate were (1) those who did not wish to be followed-up­

2 Kasturiratne A, et al. BMJ Open 2020;10:e038772. doi:10.1136/bmjopen-2020-038772 Open access BMJ Open: first published as 10.1136/bmjopen-2020-038772 on 29 September 2020. Downloaded from

≥150 mg/dL or a high-density­ lipoprotein (HDL) choles- terol level below sex specific cut-offs­ for HDL (≤40 mg/ dL for men and ≤50 mg/dL for women) at baseline. Survival analysis was done using right censored data. Follow-­up time was defined as the time between 2007 to death, or the date at which the patient was last confirmed to be alive (censoring time). Comparison of survival rates due to ACM, CVM and CRM by age-sex­ strata within ACM, CVM and CRM were conducted using Kaplan-Meier­ survival function curves. Cox’s proportional hazards (CPH) models were used to identify independent risk factors of mortality due to ACM, CVM and CRM. Initially the probable risk factors were screened with simple CPH models and the risk factors with a p<0.2 were considered for constructing multivariable CPH models for ACM, CVM and CRM. Effects of risk factors on mortality are Figure 1 The flow of participants through the follow-­up presented as HRs and their 95% CIs. period. We standardised the observed mortality rates using the method of indirect standardisation for the district and in the study, (2) those who migrated out of the area or the country and constructed life tables for the cohort. (3) those who migrated overseas. The flow of participants through the follow-­up period and how their outcome was Missing data ascertained in given in figure 1. Among variables collected at baseline, the percentage of missing data was small. We censored our outcome vari- Data analysis ables, ACM, CVM and CRM at the last date of contact We constructed Kaplan-Meier­ survival curves to deter- with the research staff, if they could not be traced in the mine the survival probability of the cohort after community or if they were already lost to follow-up.­ We accounting for all-­cause mortality, and mortality due did not perform sensitivity analyses. We have indicated to cardiovascular and cancer-­related causes, at 10 years the numbers available for each variable in table 1. We from recruitment. The participants who were not trace- have also compared the baseline socio-demographic­ able in 2017 were censored at their last date of contact. characteristics of the participants who were and were not We compared deceased persons and persons alive by traced in online supplemental table 1. important baseline characteristics. We considered socio-­ demographic variables defined as follows: age defined Potential sources of bias http://bmjopen.bmj.com/ by number of completed years at recruitment, sex as the We have attempted to address potential sources of bias reported biological sex, low educational level as having using a set of strategies in the design of this study. The an educational level below General Certificate of Educa- participants were sampled using stratified random tion (G.C.E.) Ordinary Level completion, low income sampling allowing for representative sampling from the as belonging to the lowest income quartile of the study reference population within the specific age strata to population and occupation based on the classification address selection bias. Information bias was minimised of International Labour Office.19 Smoking status was through the use of standard interviewer-administered­ defined by reported use on an average day (current questionnaires for data collection and regular in-­service on October 1, 2021 by guest. Protected copyright. smoker/past smoker/never smoker). Harmful alcohol training of data collectors who were from a medical or use was defined by Asian cut-­offs for harmful alcohol health sciences background. Certain data collected from consumption (14 IU/week for men and 7 IU/week for participants have been verified with documented infor- women) estimated using reported weekly consumption. mation, for example, medical history. Outcome data Obesity was defined as a body mass index ≥25 kg/m2. have been obtained from official documents minimising A waist circumference >90 cm for men and >80 cm for information bias. We have also attempted to minimise women was defined as a high waist circumference. Pres- confounding through use of multivariable models to ence of diabetes was defined as a verifiable past history of determine predictors of mortality. diabetes or having a fasting blood glucose level ≥126 mg/ dL at baseline. Hypertension was defined as a verifiable Patient and public involvement past history of high blood pressure or persistent systolic The study was designed in 2007. At the design stage, blood pressure ≥140 mm Hg and/or diastolic blood pres- participants were not consulted on development of sure ≥90 mm Hg at baseline. Dyslipidaemia was defined as research questions or outcome measures. However, non-­ a verifiable past history of dyslipidaemia or having a total communicable diseases, especially cardiovascular disease, cholesterol level ≥240 mg/dL or a low-density­ lipoprotein diabetes and hypertension were recognised as important cholesterol level ≥160 mg/dL or a serum triglyceride level public health problems by the public and the study

Kasturiratne A, et al. BMJ Open 2020;10:e038772. doi:10.1136/bmjopen-2020-038772 3 Open access BMJ Open: first published as 10.1136/bmjopen-2020-038772 on 29 September 2020. Downloaded from

Table 1 Baseline characteristics of the RHS cohort Deceased Deceased from Deceased from from all cardio-­vascular cancer-­related Characteristic* Total Alive causes causes causes Mean age (SD) years (n=2985) 52.4 (7.8) 52.2 (7.8) 56.9 (6.0) 57.4 (5.4) 57.5 (5.2) Age group: 35–44 years 514 (17.2) 509 (18.1) 5 (3.0) 1 (1.4) 0 (0.0)  45–54 years 1140 (38.2) 1092 (38.8) 48 (28.4) 19 (26.0) 9 (29.0)  55–64 years 1331 (44.6) 1215 (43.1) 116 (68.6) 53 (72.6) 22 (71.0) Gender: Males (n=2985) 1349 (45.2) 1242 (44.1) 107 (63.3) 46 (63.0) 19 (61.3) Education (n=2971) Below G.C.E. Ordinary Level 1358 (45.7) 1263 (45.1) 95 (56.2) 35 (47.9) 16 (51.6) Occupation (n=2964) Administrative and professional 285 (9.6) 273 (9.8) 12 (7.1) 5 (6.9) 2 (6.4) Clerical and technical 619 (20.9) 582 (20.8) 37 (21.9) 17 (23.3) 6 (19.4) Elementary occupations 188 (6.3) 170 (6.1) 18 (10.6) 6 (8.2) 6 (19.4) Unemployed 1872 (63.2) 1770 (63.3) 102 (60.4) 45 (61.6) 17 (54.8) Monthly family income (n=2921) Lowest income quartile 426 (14.6) 385 (14.0) 41 (25.0) 17 (23.9) 8 (26.7) Smoking (n=2985)  Non-­smokers 2131 (71.4) 2045 (72.6) 86 (50.9) 39 (53.4) 16 (51.6)  Ex-­smokers 368 (12.3) 323 (11.5) 45 (26.6) 20 (27.4) 7 (22.6)  Current smokers 486 (16.3) 448 (15.9) 38 (22.5) 14 (19.2) 8 (25.8) Alcohol (n=2985) Harmful alcohol users 554 (18.6) 505 (17.9) 49 (29.0) 21 (28.8) 9 (29.0) Overweight and obese (n=2974)  BMI ≥25 kg/m2 1164 (39.1) 1105 (39.4) 59 (34.9) 29 (39.7) 13 (41.9) High waist circumference (n=2983) Men >90 cm, Women >80 cm 1618 (54.2) 1539 (54.7) 79 (46.7) 37 (50.7) 16 (51.6) http://bmjopen.bmj.com/ Diabetes (n=2983) 709 (23.8) 629 (22.4) 80 (47.4) 39 (53.4) 13 (41.9) Hypertension (n=2984) 1311 (43.9) 1215 (43.2) 96 (56.8) 49 (67.1) 18 (58.1) Dyslipidaemia (n=2969) 1909 (64.3) 1802 (64.3) 107 (63.7) 50 (68.5) 17 (56.7)

*Presented as number (percentage) unless specified otherwise. BMI, body mass index; G.C.E., General Certificate of Education; RHS, Ragama Health Study.

addressed a well-recognised­ gap in knowledge. The public 17.2, 38.2% and 44.6% of the sample were distributed on October 1, 2021 by guest. Protected copyright. was involved in participant recruitment and follow-up­ in in the three 10-­year age groups 35 to 45 years, 45 to 54 this study. Community volunteers provided support to the years and 55 to 64 years. Overall, 45.5% had received an research team to access selected participants throughout education below G.C.E. Ordinary Level and 14.3% were the recruitment phase. In the follow-up­ phase, the death in the lowest income quartile. Selected baseline socio-­ registration documents available with the family members demographic, lifestyle, anthropometric and clinical char- of the participants was the main source of outcome infor- acteristics of the participants are presented in relation mation. They were retrieved with the co-­operation of to their vital status in table 1. During the period 2007 the family members. The results of this analysis will be to 2018, 169 (107 men) participants had died giving an disseminated to the study participants and their families overall crude mortality rate of 7.9% for men and 3.8% for through the RHS clinic and information leaflets deliv- women. Standardised mortality rates were 5.3 and 2.4 per ered to their homes via post. 1000 years of follow-­up for men and women, respectively. Survival probability of the cohort overall and in relation RESULTS to cardiovascular and cancer-related­ causes are shown At recruitment, the mean age of our cohort was 52.4 in figure 2. Overall survival of the cohort at 10 years was (SD=7.8) years with 1349/2985 (45.2%) men. Respectively 95.1% (94.3 to 95.9). Survival probabilities of men and

4 Kasturiratne A, et al. BMJ Open 2020;10:e038772. doi:10.1136/bmjopen-2020-038772 Open access BMJ Open: first published as 10.1136/bmjopen-2020-038772 on 29 September 2020. Downloaded from

Figure 2 Survival probability of the cohort overall and in Figure 3 Survival probabilities of men and women in relation to cardiovascular and cancer-­related causes. ACM, baseline age groups. ACM, all-­cause mortality; CRM, cancer-­ all-­cause mortality; CRM, cancer-­related mortality; CVM, related mortality; CVM, cardiovascular mortality. cardiovascular mortality. CI: 1.60 to 4.26)) had a significantly higher risk of CVM. women in baseline age groups 35 to 44, 45 to 54 and 55 Older adults (55 to 64 years HR=3.14 (95% CI: 1.45 to to 64 years for ACM, CVM and CRM are presented in 6.83)), ex-­smokers (HR=2.64 (95% CI: 1.09 to 6.42)) and figure 3. those with diabetes (HR=2.41 (95% CI: 1.18 to 4.92)) http://bmjopen.bmj.com/ HRs of selected individual variables for ACM, CVM had a significantly higher risk of CRM. The unemployed and CRM are presented in table 2. Older adults (45 to 54 (HR=0.27 (95% CI: 0.10 to 0.68)) and clerical and tech- years HR=4.35 (95% CI: 1.73 to 10.92)), (55 to 64 years, nical workers (HR=0.29 (95% CI: 0.09 to 0.91)) had a HR=9.21 (95% CI: 3.76 to 22.5), men (HR=2.15 (95% significantly lower risk of ACM. CI: 1.58 to 2.95)), those with a lower educational level Results of multiple variable analysis performed to deter- (HR=1.55 (95% CI: 1.15 to 2.11)) and a lower income mine independent predictors of mortality are presented (HR=1.95 (95% CI: 1.37 to 2.78)), smokers (ex-­smokers in table 3. The risk of ACM increased in those between

HR=3.20 (95% CI: 2.23 to 4.60)) and (current smokers 45 and 54 years of age by 3.42 (95% CI: 1.35 to 8.63) and on October 1, 2021 by guest. Protected copyright. HR=1.97 (95% CI: 1.34 to 2.88)), harmful alcohol users in those between 55 and 64 years of age by 5.99 (95% CI: (HR=1.80 (95% CI: 1.29 to 2.51)), those with diabetes 2.42 to 14.83) as compared with those between 35 and (HR=3.03 (95% CI: 2.24 to 4.11)) and hypertension 44 years of age; in those in the lowest income quartile (HR=1.69 (95% CI: 1.25 to 2.30)) had a significantly by 1.91 (95% CI: 1.33 to 2.75) as compared with those higher risk of ACM. The unemployed (HR=0.54 (95% CI: with a higher income; in a past smoker by 2.95 (95% CI: 0.32 to 0.88)), professionals or administrators (HR=0.44 2.02 to 4.32) and a current smoker by 2.60 (95% CI: 1.75 (95% CI: 0.21 to 0.91)) and those with a higher waist to 3.87) as compared with a non-­smoker; and in those circumference (HR=0.72 (95% CI: 0.53 to 0.98)) had a with diabetes at baseline by 2.86 (95% CI: 2.08 to 3.93) as significantly lower risk of ACM. Older adults (45 to 54 compared with non-­diabetics. The risk of CVM increased years HR=8.61 (95% CI: 1.15 to 64.28)), (55 to 64 years in those between 55 and 64 years of age by 10.10 (95% HR=20.80 (95% CI: 2.88 to 150.45)), males (HR=2.12 CI: 1.37 to 74.61) as compared with those between 35 and (95% CI: 1.32 to 3.41)), those with a lower income 44 years of age; in those in the lowest income quartile by (HR=1.86 (95% CI: 1.08 to 3.22)), ex-­smokers (HR=3.20 1.83 (95% CI: 1.04 to 3.21) as compared with those with a (95% CI: 1.86 to 5.50)), harmful alcohol users (HR=1.81 higher income; in a past smoker by 2.77 (95% CI: 1.55 to (95% CI: 1.09 to 3.01)), those with diabetes (HR=3.95 4.97) and a current smoker by 2.31 (95% CI: 1.21 to 4.41) (95% CI: 2.48 to 6.27)) and hypertension (HR=2.61 (95% as compared with a non-smoker;­ in those with diabetes

Kasturiratne A, et al. BMJ Open 2020;10:e038772. doi:10.1136/bmjopen-2020-038772 5 Open access BMJ Open: first published as 10.1136/bmjopen-2020-038772 on 29 September 2020. Downloaded from

Table 2 HRs of individual level baseline characteristics for all cause, cardiovascular and cancer-r­elated morality HR* (95% CI) Characteristic All-­cause mortality Cardiovascular mortality Cancer-­related mortality Age 1.09 (1.07 to 1.12) 1.11 (1.07 to 1.15) 1.11 (1.05 to 0.18) Age group*  Reference level  Reference level  Reference level 35–44 years 45–54 years 4.35 (1.73 to 10.92) 8.61 (1.15 to 64.28) Reference level 55–64 years 9.21 (3.76 to 22.5) 20.80 (2.88 to 150.45) 3.14 (1.45 to 6.83) Gender    Females Reference level Reference level Reference level Males 2.15 (1.58 to 2.95) 2.12 (1.32 to 3.41) 1.99 (0.97 to 4.10) Education    Above G.C.E. Ordinary Level Reference level Reference level Reference level Below G.C.E. Ordinary Level 1.55 (1.15 to 2.11) 1.13 (0.71 to 1.80) 1.27 (0.63 to 2.57) Occupation    Elementary occupations Reference level Reference level Reference level Administrative and professional 0.44 (0.21 to 0.91) 0.54 (0.17 to 1.78) 0.22 (0.04 to 1.10) Clerical and technical 0.61 (0.34 to 1.06) 0.84 (0.33 to 2.13) 0.29 (0.09 to 0.91) Unemployed 0.54 (0.32 to 0.88) 0.70 (0.30 to 1.65) 0.27 (0.10 to 0.68) Monthly family income    Upper income quartiles Reference level Reference level Reference level Lowest income quartile 1.95 (1.37 to 2.78) 1.86 (1.08 to 3.22) 2.08 (0.93 to 4.69) Smoking     Non-­smoker Reference level Reference level Reference level  Ex-­smoker 3.20 (2.23 to 4.60) 3.20 (1.86 to 5.50) 2.64 (1.09 to 6.42)  Current smoker 1.97 (1.34 to 2.88) 1.62 (0.88 to 2.99) 2.19 (0.94 to 5.12) Alcohol    Non-harmful­ alcohol user Reference level Reference level Reference level

Harmful alcohol user 1.80 (1.29 to 2.51) 1.81 (1.09 to 3.01) 1.76 (0.81 to 3.83) http://bmjopen.bmj.com/ BMI     BMI <25 kg/m2 Reference level Reference level Reference level  BMI ≥25 kg/m2 0.81 (0.59 to 1.11) 0.98 (0.61 to 1.58) 1.10 (0.54 to 2.25) Waist circumference    Males ≤90 cm, Females ≤80 cm Reference level Reference level Reference level Males >90 cm, Females >80 cm 0.72 (0.53 to 0.98) 0.84 (0.53 to 1.34) 0.88 (0.43 to 1.78)

Diabetes 3.03 (2.24 to 4.11) 3.95 (2.48 to 6.27) 2.41 (1.18 to 4.92) on October 1, 2021 by guest. Protected copyright. Hypertension 1.69 (1.25 to 2.30) 2.61 (1.60 to 4.26) 1.80 (0.88 to 3.67) Dyslipidaemia 0.97 (0.71 to 1.33) 1.19 (0.73 to 1.96) 0.73 (0.35 to 1.50)

Bold statistics indicate significant variables - that is, 1 is not included in the 95% CI. *The age groups 35 to 44 years, 45 to 54 years and 55 to 64 years have been sampled in 1:2:2 ratio BMI, body mass index; G.C.E., General Certificate of Education.

at baseline by 3.30 (95% CI: 2.07 to 5.58) and those with a lower risk of CRM as compared with those in elemen- hypertension at baseline by 1.90 (95% CI: 1.12 to 3.25) as tary occupations. Indirect standardisation of observed compared with those without. The risk of CRM increased mortality rates for men and women showed that mortality in those between 55 and 64 years of age by 3.19 (95% in this cohort was about 50% lower compared with the CI: 1.44 to 7.05) as compared with those between 35 and district and national populations (table 4). 54 years of age. Those in clerical and technical occupa- Comparison of basic socio-­demographic characteris- tions (HR=0.31 (95% CI: 0.10 to 0.96)) and those who tics of the RHS cohort who were traced and not traced are unemployed (HR=0.21 (95% CI: 0.08 to 0.55)) had in 2017/2018 for outcome ascertainment is presented in

6 Kasturiratne A, et al. BMJ Open 2020;10:e038772. doi:10.1136/bmjopen-2020-038772 Open access BMJ Open: first published as 10.1136/bmjopen-2020-038772 on 29 September 2020. Downloaded from

Table 3 Risk factors of all-­cause, cardiovascular and cancer mortality from multiple variable analysis Risk factor (at baseline) HR 95% CI P value All-­cause mortality Age group : 45–54 vs 35–44 years 3.42 1.35 to 8.63 0.009 55–64 vs 35–44 years 5.99 2.42 to 14.83 <0.001 Lowest income quartile 1.91 1.33 to 2.75 <0.001 Smoking : Ex-­smokers vs non-­smoker 2.95 2.02 to 4.32 <0.001  Smokers vs non-­smoker 2.60 1.75 to 3.87 <0.001 Diabetes 2.86 2.08 to 3.93 <0.001 Cardiovascular mortality Age group : 45–54 vs 35–44 years 5.54 0.73 to 41.81 0.097 55–64 vs 35–44 years 10.10 1.37 to 74.61 0.023 Lowest income quartile 1.83 1.04 to 3.21 0.035 Smoking : Ex-­smokers vs non-­smoker 2.77 1.55 to 4.97 <0.001  Smokers vs non-­smoker 2.31 1.21 to 4.41 0.011 Diabetes 3.30 2.07 to 5.58 <0.001 Hypertension 1.90 1.12 to 3.25 0.018 Cancer-­related mortality Age group : 55–64 vs 35–54 years* 3.19 1.44 to 7.05 0.004 Occupation: Professional/administration vs elementary 0.22 0.04 to 1.07 0.061 Clerical/technical vs elementary 0.31 0.10 to 0.96 0.042 Unemployed vs elementary 0.21 0.08 to 0.55 0.001

*There were no cancer-­related deaths in 35 to 44 age group. Therefore, age groups 35 to 44 and 45 to 54 were merged together. online supplemental table 1. Life tables constructed for identified as the third leading contributor for deaths in the RHS cohort are presented in online supplemental Sri Lanka, behind ischaemic heart disease and stroke.23 table 2. Similar to documented findings,24–26 both current and past smoking was associated with ACM. Hypertension at baseline was associated with CVM, similar to findings http://bmjopen.bmj.com/ DISCUSSION from a number of cohorts across settings in both high-­ We have demonstrated the patterns and predictors of income and middle-income­ regions.12 20 Our findings ACM, CVM and CRM in a cohort of semi-­urban adults highlight the importance of well-known­ non-­modifiable living in Sri Lanka, using baseline data of a prospec- predictors of mortality, such as, older age, while at the tive population-based­ cohort study and mortality data same time demonstrating that well-recognised­ modifi- from the national vital registration system. Our cohort able risk factors of mortality like low income, elementary comprised middle-­aged adults with a mean age of 52.4 occupations, current and past smoking, diabetes and (SD=7.8) years at baseline and 45% men. The overall on October 1, 2021 by guest. Protected copyright. hypertension are associated with ACM and CVM in this survival probability of the cohort was 95.1% at 10 years population. after recruitment. ACM was associated with older age groups, low income, both current and previous smoking Standardisation of the mortality data of our cohort for and diabetes at baseline. CVM was associated with older the district showed that our population is healthier than age groups, low income, both current and previous the general population. The RHS cohort population lives smoking, diabetes and hypertension at baseline. CRM in the catchment area of the Colombo North Teaching was associated with older age and those engaged in cler- Hospital which is the second largest tertiary care hospital ical and technical occupations and those who are unem- in the country in terms of bed strength and healthcare ployed had a lower risk of CRM as compared with those in facilities. The area also has two large private hospitals elementary occupations. and many general practice clinics. In addition, the popu- Older age is the main risk factor for mortality glob- lation has exclusive access to long-­term and continuing ally.20 We found that low income was associated with both health facilities in the RHS clinic at the Faculty of Medi- ACM and CVM, while CRM was associated with elemen- cine, University of Kelaniya, since the formation of the tary occupations. The association between poverty and cohort to-date.­ This service has offered routine follow-­up mortality is well documented.21 22 Diabetes was also care, health advice and a continuous supply of medica- associated with both ACM and CVM. Diabetes has been tion to all participants who have metabolic conditions

Kasturiratne A, et al. BMJ Open 2020;10:e038772. doi:10.1136/bmjopen-2020-038772 7 Open access BMJ Open: first published as 10.1136/bmjopen-2020-038772 on 29 September 2020. Downloaded from

Table 4 Indirect standardisation of mortality for men and women Total number of person Expected number of Expected number of years within the age Total number of deaths if applied to deaths if applied to Age group (years) group observed deaths Gampaha district Sri Lanka Men 35–39 237 0 0 1 40–44 839 0 2 3 45–49 1555 3 8 8 50–54 2168 3 17 17 55–59 2657 20 31 31 60–64 2790 22 47 49 65–69 2090 40 55 58 70–74 763 18 31 33 75+ 36 1 2 2 Total 13 135 107 193 202 Women 35–39 298 0 0 0 40–44 962 0 1 1 45–49 1870 2 3 4 50–54 2765 4 8 9 55–59 3422 6 16 17 60–64 3420 18 25 27 65–69 2566 20 34 37 70–74 1014 12 24 26 75+ 45 0 2 2 Total 16 362 62 113 123 like diabetes and hypertension with referral links to benefits and property settlements, and are retained by the specialised care based on requirement. Volunteer bias in next-­of-­kin of the deceased and the local public adminis- http://bmjopen.bmj.com/ participation in screening and research programmes and tration system.29 In the absence of complete and up-to-­ ­ intensified health services available to the participants of date health information systems covering risk levels and this cohort could be responsible for their apparent better outcome information, the methods we have employed, health status. which are straightforward and simple enough to be easily Our cohort of 2985 participants is a relatively small cohort replicated by other researchers with similar data sources, from South Asia. However, our comprehensive baseline can be used to determine the patterns and predictors of data and stringent follow-up­ mechanisms, together with a mortality in any low-income­ and middle-­income setting reliable vital registration system strengthen the relevance with a reasonable vital registration system. on October 1, 2021 by guest. Protected copyright. of our findings and their applicability for the country and The predictors of mortality reported in this study are the region. Population based cohort studies provide the well known. However, this is the first time this has been most robust methodology for determining the temporal reported for this setting. The added benefits of reporting association between a risk factor and its outcome in a these results are (1) it can increase the emphasis on population.27 We have taken steps to address the effect preventing identified modifiable risks, (2) allocation of of confounding on these relationships through multiple limited health resources can be planned to target these variable analysis. Sri Lanka’s vital registration system has conditions and (3) increase the attention of policymakers been in operation for over 100 years.28 Transformations to prevent premature mortality. in knowledge and technology over the last two to three decades have led to improvement in documentation of Limitations vital events and the use of modern information tech- The relatively small size of the cohort and the modest nology for collecting, storage and retrieval of vital statis- number of mortality outcomes reported over the first 10 tics.11 It is a legal requirement for all births and deaths years of follow-up­ is our main limitation. Due to the small to be registered in the vital registration system.28 Further- number of outcomes we were compelled to categorise more, availability of the death certificate issued by the mortality into three main groups as ACM, CVM and CRM. Registrar General’s Department is mandatory for death We expect to follow-up­ this cohort long-term­ and examine

8 Kasturiratne A, et al. BMJ Open 2020;10:e038772. doi:10.1136/bmjopen-2020-038772 Open access BMJ Open: first published as 10.1136/bmjopen-2020-038772 on 29 September 2020. Downloaded from the mortality trends over the next 10 years to address this Competing interests None declared. limitation. Despite improvements in recording cause of Patient and public involvement Patients and/or the public were involved in the death in mortality statistics, deficiencies can exist in the design, or conduct, or reporting, or dissemination plans of this research. Refer to death registration process that can reduce the accuracy the Methods section for further details. of documenting the cause of death. Validating the docu- Patient consent for publication Not required. mented cause of death in official documents was beyond Ethics approval Ethical clearance was obtained from the Ethics Review the scope of this study and we acknowledge this as a Committee, Faculty of Medicine, University of Kelaniya, for both baseline and follow-­up assessments of the Ragama Health Study and principles of ethical potential limitation. research were upheld in all encounters with the participants. Sri Lanka being a small country with access to free Provenance and peer review Not commissioned; externally peer reviewed. education and healthcare has minimal country-wide­ varia- Data availability statement Data are available upon reasonable request. De-­ tion. Given the diversity of the study area in the Gampaha identified participant data will be available upon reasonable request as decided district, we believe that the risk factor prevalences and by the data management committee of the RHS. Please contact the corresponding associations reported in different subpopulations in this author (Anuradhani Kasturiratne,ORCID ID 0000-0001-5260-2394). study will be the same for generalising the findings of the Open access This is an open access article distributed in accordance with the study both in Sri Lanka and the South Asian region. Creative Commons Attribution Non Commercial (CC BY-­NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially­ , and license their derivative works on different terms, provided the original work is Recommendations properly cited, appropriate credit is given, any changes made indicated, and the use Low income is known to be associated with increased is non-­commercial. See: http://​creativecommons.org/​ ​licenses/by-​ ​nc/4.​ ​0/. ACM across high-income,­ middle-income­ and low-income­ ORCID iDs countries and even where health systems are tax funded Anuradhani Kasturiratne http://orcid.​ ​org/0000-​ ​0001-5260-​ ​2394 and provided free at the point of delivery.22 30 Many Arunasalam Pathmeswaran http://orcid.​ ​org/0000-​ ​0003-4065-​ ​2639 underlying mechanisms that link poverty with poorer Udaya Ranawaka http://orcid.​ ​org/0000-​ ​0002-4050-​ ​062X health outcomes like socio-economic­ barriers for access to healthy behaviours, and the resulting choice of unhealthy 31 32 alternatives, exist at population level. Universal REFERENCES measures to improve living standards by addressing 1 Bonneux L. How to measure the burden of mortality? J Epidemiol Community Health 2002;56:128–31. poverty can create manifold benefits including the 2 United Nations, Department of Economic and Social Affairs PD. promotion of healthy behaviours that can reduce the World mortality 2017 data booklet, 2017. Available: https://www.​ incidence and improve the control of intermediate meta- un.​org/​en/​development/​desa/​population/​publications/​pdf/​mortality/​ World-​Mortality-​2017-​Data-​Booklet.​pdf bolic risk factors of common mortality outcomes, and 3 GBD 2017 Causes of Death Collaborators. Global, regional, and thus improve longevity. Along with this, preventive health national age-­sex-­specific mortality for 282 causes of death in 195 countries and territories, 1980-2017: a systematic analysis for the services should increase their emphasis on behavioural global burden of disease study 2017. Lancet 2018;392:1736–88. modification for healthy lifestyles at population level to 4 Unnikrishnan R, Gupta PK, Mohan V. Diabetes in South Asians: address tobacco control, and prevention and control of phenotype, clinical presentation, and natural history. Curr Diab Rep http://bmjopen.bmj.com/ 2018;18:30. diabetes and hypertension. 5 Lee JR, Yeh H-C.­ Trends in the prevalence of type 2 diabetes and its association with mortality rates in Asians vs. whites: results from the Author affiliations United States National health interview survey from 2000 to 2014. J 1 Diabetes Complications 2018;32:539–44. Department of Public Health, University of Kelaniya Faculty of Medicine, Ragama, 6 Banerjee AT, Shah BR. Differences in prevalence of diabetes among Sri Lanka immigrants to Canada from South Asian countries. Diabet Med 2Centre for Health Informatics, Biostatistics and Epidemiology, University of Kelaniya 2018;35:937–43. Faculty of Medicine, Ragama, Sri Lanka 7 Cancer IA for Ron. WHO south-­east Asia region, 2018. Available: 3Department of Medicine, University of Kelaniya Faculty of Medicine, Ragama, Sri https://​gco.​iarc.​fr/​today/​data/​factsheets/​populations/​995-​who-​ south-​east-​asia-​region-​searo-​fact-​sheets.​pdf [Accessed 14 Jun Lanka on October 1, 2021 by guest. Protected copyright. 4 2020]. Department of Pharmacology, University of Kelaniya Faculty of Medicine, Ragama, 8 World Health Organization. Global health estimates, 2016. Sri Lanka 9 Vathesatogkit P, Batty GD, Woodward M. Socioeconomic 5Department of Biochemistry and Clinical Chemistry, University of Kelaniya Faculty disadvantage and disease-­specific mortality in Asia: systematic of Medicine, Ragama, Sri Lanka review with meta-­analysis of population-­based cohort studies. J 6Department of Gene Diagnostics and Therapeutics, National Center for Global Epidemiol Community Health 2014;68:375–83. 10 Anjana RM, Unnikrishnan R, Mugilan P, et al. Causes and predictors Health and Medicine Research Institute, Shinjuku-­ku, Tokyo, Japan of mortality in Asian Indians with and without diabetes-10 year follow-­up of the Chennai Urban Rural Epidemiology Study (CURES - Contributors AK, NK, ARW and HJdS conceptualised and designed the study. 150). PLoS One 2018;13:e0197376. STDS, MAN, ASD, APDS, SC and UR managed the study implementation and 11 Registrar General’s Department SL. Death registration, 2019. data collection. AP, DSE and UBT managed the database and conducted the data Available: http://www.​rgd.​gov.​lk/​web/​index.​php/​en/ [Accessed 5 Nov 2019]. analysis. AK and HJdS drafted the manuscript. All authors contributed to improving 12 Coates MM, Kamanda M, Kintu A, et al. A comparison of all-­ and approval of the final version of the manuscript. cause and cause-­specific mortality by household socioeconomic Funding This work was supported by grants from the National Center for Global status across seven indepth network health and demographic Health and Medicine (17K1); KAKENHI (Grant-­in-­Aid for Scientific Research (B) surveillance systems in sub-Saharan­ Africa. Glob Health Action 2019;12:1608013–12. #20406019) from the Ministry of Education, Culture, Sports, Science and Technology 13 Foreman KJ, Marquez N, Dolgert A, et al. Forecasting life of Japan; Japan International Co-­operation Agency through the project on Health expectancy, years of life lost, and all-cause­ and cause-specific­ Promotion and Preventive Care Measures of Chronic Non-communicable­ Diseases mortality for 250 causes of death: reference and alternative (NPP) implemented by the Ministry of Health, Sri Lanka (project #0701475); scenarios for 2016-40 for 195 countries and territories. Lancet Strengthening Research Grant from the Ministry of Higher Education, Sri Lanka. 2018;392:2052–90.

Kasturiratne A, et al. BMJ Open 2020;10:e038772. doi:10.1136/bmjopen-2020-038772 9 Open access BMJ Open: first published as 10.1136/bmjopen-2020-038772 on 29 September 2020. Downloaded from

14 Department of Census and Statistics Sri Lanka. Census of 24 Nakamura K, Huxley R, Ansary-Moghaddam­ A, et al. The hazards population and housing 2012 Sri Lanka, 2015. Available: http:// and benefits associated with smoking and smoking cessation in Asia: www.​statistics.​gov.​lk/​PopHouSat/​CPH2011/​index.​php?​fileName=​ a meta-analysis­ of prospective studies. Tob Control 2009;18:345–53. Activities/​Tent​ativ​elis​tofP​ubli​cations [Accessed 1 Nov 2019]. 25 Duong M, Rangarajan S, Zhang X, et al. Effects of bidi smoking on 15 Election Commission of Sri Lanka. Universal franchise and elections, all-cause­ mortality and cardiorespiratory outcomes in men from 2019. Available: https://​elections.​gov.​lk/​web/​en/​about-​us/​what-​we-​ South Asia: an observational community-­based substudy of the do/​universal-​franchise-​and-​election/ prospective urban rural epidemiology study (pure). Lancet Glob 16 Ministry of Health SL. Annual health statistics, 2017. Available: http:// Health 2017;5:e168–76. www.​moh.​gov.​sa/​en/​Ministry/​Statistics/​book/​Pages/​default.​aspx 26 GBD 2015 Tobacco Collaborators. Smoking prevalence and 17 Ministry of Health N and IM. Mid year population estimates for attributable disease burden in 195 countries and territories, 1990- medical officer of health (MOH) areas 2018. Colombo, 2018. 2015: a systematic analysis from the global burden of disease study 18 Dassanayake AS, Kasturiratne A, Rajindrajith S, et al. Prevalence 2015. Lancet 2017;389:1885–906. and risk factors for non-alcoholic­ fatty liver disease among adults 27 Grimes DA, Schulz KF. Cohort studies: marching towards outcomes. in an urban Sri Lankan population. J Gastroenterol Hepatol Lancet 2002;359:341–5. 2009;24:1284–8. 28 Rampatige R, Gamage S, Peiris S, et al. Assessing the reliability of 19 Office IL. International standard classification of occupations: ISCO- causes of death reported by the vital registration system in Sri Lanka: 08, 2012. medical records review in Colombo. Health Inf Manag 2013;42:20–8. 20 Preis SR, Hwang S-­J, Coady S, et al. Trends in all-cause­ and 29 Democratic Socialist Republic of Sri Lanka. Births and deaths cardiovascular disease mortality among women and men with and registration act Amendment. Sri Jayawardenepura Kotte: Parliment without diabetes mellitus in the Framingham heart study, 1950 to of Sri Lanka, 2005. 2005. Circulation 2009;119:1728–35. 30 Shin WY, Kim HC, Lee T, et al. Combined effects of diabetes and low 21 Lynch JW, Smith GD, Kaplan GA, et al. Income inequality and household income on mortality: a 12-year­ follow-up­ study of 505 677 mortality: importance to health of individual income, psychosocial Korean adults. Diabet Med 2018;35:1345–54. environment, or material conditions. BMJ 2000;320:1200–4. 31 Miller V, Yusuf S, Chow CK, et al. Availability, affordability, and 22 Kinge JM, Modalsli JH, Øverland S, et al. Association of household consumption of fruits and vegetables in 18 countries across income income with life expectancy and cause-­specific mortality in Norway, levels: findings from the prospective urban rural epidemiology (pure) 2005-2015. JAMA 2019;321:1916–25. study. Lancet Glob Health 2016;4:e695–703. 23 Institute of Health Metrics and Evaluation. GBD 2017, 2019. 32 Mayén A-L,­ Marques-Vidal­ P, Paccaud F, et al. Socioeconomic Available: https://​vizhub.​healthdata.​org/​gbd-​compare/ [Accessed 5 determinants of dietary patterns in low- and middle-­income Nov 2019]. countries: a systematic review. Am J Clin Nutr 2014;100:1520–31. http://bmjopen.bmj.com/ on October 1, 2021 by guest. Protected copyright.

10 Kasturiratne A, et al. BMJ Open 2020;10:e038772. doi:10.1136/bmjopen-2020-038772